Scater/Scran:: Spatial Transcriptomics
Åsa Björklund & Paulo Czarnewski
January 14, 2022
Spatial transcriptomics
Spatial transcriptomic data with the Visium platform is in many ways similar to scRNAseq data. It contains UMI counts for 5-20 cells instead of single cells, but is still quite sparse in the same way as scRNAseq data is, but with the additional information about spatial location in the tissue.
Here we will first run quality control in a similar manner to scRNAseq data, then QC filtering, dimensionality reduction, integration and clustering. Then we will use scRNAseq data from mouse cortex to run LabelTransfer to predict celltypes in the Visium spots.
We will use two Visium spatial transcriptomics dataset of the mouse brain (Sagittal), which are publicly available from the 10x genomics website. Note, that these dataset have already been filtered for spots that does not overlap with the tissue.
Load packages
mkdir -p data/visium/Posterior
mkdir -p data/visium/Anterior
cd data/visium/Posterior
curl -o V1_Mouse_Brain_Sagittal_Posterior_filtered_feature_bc_matrix.tar.gz https://cf.10xgenomics.com/samples/spatial-exp/1.1.0/V1_Mouse_Brain_Sagittal_Posterior/V1_Mouse_Brain_Sagittal_Posterior_filtered_feature_bc_matrix.tar.gz
tar xvzf V1_Mouse_Brain_Sagittal_Posterior_filtered_feature_bc_matrix.tar.gz
curl -o V1_Mouse_Brain_Sagittal_Posterior_spatial.tar.gz https://cf.10xgenomics.com/samples/spatial-exp/1.1.0/V1_Mouse_Brain_Sagittal_Posterior/V1_Mouse_Brain_Sagittal_Posterior_spatial.tar.gz
tar xvzf V1_Mouse_Brain_Sagittal_Posterior_spatial.tar.gz
rm *.tar.gz
cd ../Anterior
curl -o V1_Mouse_Brain_Sagittal_Anterior_filtered_feature_bc_matrix.tar.gz https://cf.10xgenomics.com/samples/spatial-exp/1.1.0/V1_Mouse_Brain_Sagittal_Anterior/V1_Mouse_Brain_Sagittal_Anterior_filtered_feature_bc_matrix.tar.gz
tar xvzf V1_Mouse_Brain_Sagittal_Anterior_filtered_feature_bc_matrix.tar.gz
curl -o V1_Mouse_Brain_Sagittal_Anterior_spatial.tar.gz https://cf.10xgenomics.com/samples/spatial-exp/1.1.0/V1_Mouse_Brain_Sagittal_Anterior/V1_Mouse_Brain_Sagittal_Anterior_spatial.tar.gz
tar xvzf V1_Mouse_Brain_Sagittal_Anterior_spatial.tar.gz
rm *.tar.gz
cd ..## % Total % Received % Xferd Average Speed Time Time Time Current
## Dload Upload Total Spent Left Speed
##
0 0 0 0 0 0 0 0 --:--:-- --:--:-- --:--:-- 0
3 54.8M 3 2141k 0 0 2288k 0 0:00:24 --:--:-- 0:00:24 2288k
16 54.8M 16 9405k 0 0 4864k 0 0:00:11 0:00:01 0:00:10 4863k
32 54.8M 32 18.0M 0 0 6315k 0 0:00:08 0:00:02 0:00:06 6314k
46 54.8M 46 25.5M 0 0 6650k 0 0:00:08 0:00:03 0:00:05 6649k
57 54.8M 57 31.4M 0 0 6524k 0 0:00:08 0:00:04 0:00:04 6524k
76 54.8M 76 41.9M 0 0 7205k 0 0:00:07 0:00:05 0:00:02 8122k
92 54.8M 92 50.9M 0 0 7527k 0 0:00:07 0:00:06 0:00:01 8556k
100 54.8M 100 54.8M 0 0 7661k 0 0:00:07 0:00:07 --:--:-- 8560k
## x filtered_feature_bc_matrix/
## x filtered_feature_bc_matrix/barcodes.tsv.gz
## x filtered_feature_bc_matrix/matrix.mtx.gz
## x filtered_feature_bc_matrix/features.tsv.gz
## % Total % Received % Xferd Average Speed Time Time Time Current
## Dload Upload Total Spent Left Speed
##
0 0 0 0 0 0 0 0 --:--:-- --:--:-- --:--:-- 0
0 9480k 0 6363 0 0 7255 0 0:22:18 --:--:-- 0:22:18 7255
64 9480k 64 6093k 0 0 4034k 0 0:00:02 0:00:01 0:00:01 4032k
100 9480k 100 9480k 0 0 5003k 0 0:00:01 0:00:01 --:--:-- 5003k
## x spatial/
## x spatial/tissue_positions_list.csv
## x spatial/tissue_hires_image.png
## x spatial/scalefactors_json.json
## x spatial/aligned_fiducials.jpg
## x spatial/detected_tissue_image.jpg
## x spatial/tissue_lowres_image.png
## % Total % Received % Xferd Average Speed Time Time Time Current
## Dload Upload Total Spent Left Speed
##
0 0 0 0 0 0 0 0 --:--:-- --:--:-- --:--:-- 0
4 54.8M 4 2669k 0 0 4767k 0 0:00:11 --:--:-- 0:00:11 4767k
26 54.8M 26 14.3M 0 0 9398k 0 0:00:05 0:00:01 0:00:04 9393k
51 54.8M 51 28.4M 0 0 10.8M 0 0:00:05 0:00:02 0:00:03 10.8M
80 54.8M 80 44.2M 0 0 12.4M 0 0:00:04 0:00:03 0:00:01 12.4M
100 54.8M 100 54.8M 0 0 13.1M 0 0:00:04 0:00:04 --:--:-- 13.1M
## x filtered_feature_bc_matrix/
## x filtered_feature_bc_matrix/barcodes.tsv.gz
## x filtered_feature_bc_matrix/features.tsv.gz
## x filtered_feature_bc_matrix/matrix.mtx.gz
## % Total % Received % Xferd Average Speed Time Time Time Current
## Dload Upload Total Spent Left Speed
##
0 0 0 0 0 0 0 0 --:--:-- --:--:-- --:--:-- 0
2 9017k 2 223k 0 0 711k 0 0:00:12 --:--:-- 0:00:12 710k
100 9017k 100 9017k 0 0 9811k 0 --:--:-- --:--:-- --:--:-- 9801k
## x spatial/
## x spatial/tissue_positions_list.csv
## x spatial/tissue_hires_image.png
## x spatial/scalefactors_json.json
## x spatial/detected_tissue_image.jpg
## x spatial/tissue_lowres_image.png
## x spatial/aligned_fiducials.jpg# BiocManager::install('DropletUtils',update = F)
devtools::install_github("RachelQueen1/Spaniel", ref = "Development", upgrade = F,
dependencies = F)##
checking for file ‘/private/var/folders/1s/j9ck5c_162s487xcprlxtmdh0000gp/T/RtmpVgCm2g/remotesf57f685ebf1/RachelQueen1-Spaniel-bb1bb99/DESCRIPTION’ ...
✔ checking for file ‘/private/var/folders/1s/j9ck5c_162s487xcprlxtmdh0000gp/T/RtmpVgCm2g/remotesf57f685ebf1/RachelQueen1-Spaniel-bb1bb99/DESCRIPTION’
##
─ preparing ‘Spaniel’:
##
checking DESCRIPTION meta-information ...
✔ checking DESCRIPTION meta-information
##
─ checking for LF line-endings in source and make files and shell scripts
##
─ checking for empty or unneeded directories
##
Omitted ‘LazyData’ from DESCRIPTION
##
─ building ‘Spaniel_1.2.0.tar.gz’
##
## library(Spaniel)
library(biomaRt)
suppressPackageStartupMessages(require(Matrix))
suppressPackageStartupMessages(require(dplyr))
suppressPackageStartupMessages(require(scran))
suppressPackageStartupMessages(require(SingleR))
suppressPackageStartupMessages(require(scater))
suppressPackageStartupMessages(require(ggplot2))
suppressPackageStartupMessages(require(patchwork))
suppressPackageStartupMessages(require(cowplot))Load ST data
We can first load and merge the objects into one SCE object.
sce.a <- Spaniel::createVisiumSCE(tenXDir = "data/visium/Anterior", resolution = "Low")
sce.p <- Spaniel::createVisiumSCE(tenXDir = "data/visium/Posterior", resolution = "Low")
sce <- cbind(sce.a, sce.p)
sce$Sample <- sub(".*[/]", "", sub("/filtered_feature_bc_matrix", "", sce$Sample))
lll <- list(sce.a, sce.p)
lll <- lapply(lll, function(x) x@metadata)
names(lll) <- c("Anterior", "Posterior")
sce@metadata <- lllWe can further convert the gene ensembl IDs to gene names.
mart <- biomaRt::useMart(biomart = "ENSEMBL_MART_ENSEMBL", dataset = "mmusculus_gene_ensembl")
annot <- biomaRt::getBM(attributes = c("ensembl_gene_id", "external_gene_name", "gene_biotype"),
mart = mart, useCache = F)
gene_names <- as.character(annot[match(rownames(sce), annot[, "ensembl_gene_id"]),
"external_gene_name"])
gene_names[is.na(gene_names)] <- ""
sce <- sce[gene_names != "", ]
rownames(sce) <- gene_names[gene_names != ""]
dim(sce)## [1] 32146 6050Quality control
Similar to scRNAseq we use statistics on number of counts, number of features and percent mitochondria for quality control.
Now the counts and feature counts are calculated on the Spatial assay, so they are named “nCount_Spatial” and “nFeature_Spatial”.
# Mitochondrial genes
mito_genes <- rownames(sce)[grep("^mt-", rownames(sce))]
# Ribosomal genes
ribo_genes <- rownames(sce)[grep("^Rp[sl]", rownames(sce))]
# Hemoglobin genes - includes all genes starting with HB except HBP.
hb_genes <- rownames(sce)[grep("^Hb[^(p)]", rownames(sce))]
sce <- addPerCellQC(sce, flatten = T, subsets = list(mt = mito_genes, hb = hb_genes,
ribo = ribo_genes))
head(colData(sce))## DataFrame with 6 rows and 24 columns
## Sample Barcode Section Spot_Y Spot_X Image_Y
## <character> <character> <integer> <integer> <integer> <integer>
## 1 Anterior AAACAAGTATCTCCCA-1 1 50 102 7474
## 2 Anterior AAACACCAATAACTGC-1 1 59 19 8552
## 3 Anterior AAACAGAGCGACTCCT-1 1 14 94 3163
## 4 Anterior AAACAGCTTTCAGAAG-1 1 43 9 6636
## 5 Anterior AAACAGGGTCTATATT-1 1 47 13 7115
## 6 Anterior AAACATGGTGAGAGGA-1 1 62 0 8912
## Image_X pixel_x pixel_y sum detected total sum
## <integer> <numeric> <numeric> <numeric> <integer> <numeric> <numeric>
## 1 8500 438.898 214.079 13991 4462 13991 13959
## 2 2788 143.959 158.417 39797 8126 39797 39740
## 3 7950 410.499 436.678 29951 6526 29951 29905
## 4 2100 108.434 257.349 42333 8190 42333 42262
## 5 2375 122.633 232.616 35700 8090 35700 35659
## 6 1480 76.420 139.828 22148 6518 22148 22098
## detected subsets_mt_sum subsets_mt_detected subsets_mt_percent
## <integer> <numeric> <integer> <numeric>
## 1 4457 1521 12 10.89620
## 2 8116 3977 12 10.00755
## 3 6520 4265 12 14.26183
## 4 8181 2870 12 6.79097
## 5 8082 1831 13 5.13475
## 6 6511 2390 12 10.81546
## subsets_hb_sum subsets_hb_detected subsets_hb_percent subsets_ribo_sum
## <numeric> <integer> <numeric> <numeric>
## 1 60 4 0.429830 826
## 2 831 6 2.091092 2199
## 3 111 5 0.371175 1663
## 4 117 5 0.276844 3129
## 5 73 5 0.204717 2653
## 6 134 5 0.606390 1478
## subsets_ribo_detected subsets_ribo_percent total
## <integer> <numeric> <numeric>
## 1 85 5.91733 13959
## 2 89 5.53347 39740
## 3 88 5.56094 29905
## 4 88 7.40381 42262
## 5 90 7.43992 35659
## 6 84 6.68839 22098plot_grid(plotColData(sce, y = "detected", x = "Sample", colour_by = "Sample"), plotColData(sce,
y = "total", x = "Sample", colour_by = "Sample"), plotColData(sce, y = "subsets_mt_percent",
x = "Sample", colour_by = "Sample"), plotColData(sce, y = "subsets_ribo_percent",
x = "Sample", colour_by = "Sample"), plotColData(sce, y = "subsets_hb_percent",
x = "Sample", colour_by = "Sample"), ncol = 3)
We can also plot the same data onto the tissue section.
samples <- c("Anterior", "Posterior")
to_plot <- c("detected", "total", "subsets_mt_percent", "subsets_ribo_percent", "subsets_hb_percent")
plist <- list()
n = 1
for (j in to_plot) {
for (i in samples) {
temp <- sce[, sce$Sample == i]
temp@metadata <- temp@metadata[[i]]
plist[[n]] <- spanielPlot(object = temp, plotType = "Cluster", clusterRes = j,
customTitle = j, techType = "Visium", ptSizeMax = 1, ptSizeMin = 0.1)
n <- n + 1
}
}
plot_grid(ncol = 2, plotlist = plist)
As you can see, the spots with low number of counts/features and high mitochondrial content is mainly towards the edges of the tissue. It is quite likely that these regions are damaged tissue. You may also see regions within a tissue with low quality if you have tears or folds in your section.
But remember, for some tissue types, the amount of genes expressed and proportion mitochondria may also be a biological features, so bear in mind what tissue you are working on and what these features mean.
Filter
Select all spots with less than 25% mitocondrial reads, less than 20% hb-reads and 1000 detected genes. You must judge for yourself based on your knowledge of the tissue what are appropriate filtering criteria for your dataset.
sce <- sce[, sce$detected > 500 & sce$subsets_mt_percent < 25 & sce$subsets_hb_percent <
20]
dim(sce)## [1] 32146 5803And replot onto tissue section:
samples <- c("Anterior", "Posterior")
to_plot <- c("detected", "total", "subsets_mt_percent", "subsets_mt_percent", "subsets_hb_percent")
plist <- list()
n = 1
for (j in to_plot) {
for (i in samples) {
temp <- sce[, sce$Sample == i]
temp@metadata <- temp@metadata[[i]]
plist[[n]] <- spanielPlot(object = temp, plotType = "Cluster", clusterRes = j,
customTitle = j, techType = "Visium", ptSizeMax = 1, ptSizeMin = 0.1)
n <- n + 1
}
}
plot_grid(ncol = 2, plotlist = plist)
Top expressed genes
As for scRNAseq data, we will look at what the top expressed genes are.
C = counts(sce)
C@x = C@x/rep.int(colSums(C), diff(C@p))
most_expressed <- order(Matrix::rowSums(C), decreasing = T)[20:1]
boxplot(as.matrix(t(C[most_expressed, ])), cex = 0.1, las = 1, xlab = "% total count per cell",
col = (scales::hue_pal())(20)[20:1], horizontal = TRUE)
rm(C)As you can see, the mitochondrial genes are among the top expressed. Also the lncRNA gene Bc1 (brain cytoplasmic RNA 1). Also one hemoglobin gene.
Filter genes
We will remove the Bc1 gene, hemoglobin genes (blood contamination) and the mitochondrial genes.
dim(sce)## [1] 32146 5803# Filter Bl1
sce <- sce[!grepl("Bc1", rownames(sce)), ]
# Filter Mitocondrial
sce <- sce[!grepl("^mt-", rownames(sce)), ]
# Filter Hemoglobin gene (optional if that is a problem on your data)
sce <- sce[!grepl("^Hb.*-", rownames(sce)), ]
dim(sce)## [1] 32124 5803Analysis
sce <- computeSumFactors(sce, sizes = c(20, 40, 60, 80))
sce <- logNormCounts(sce)Now we can plot gene expression of individual genes, the gene Hpca is a strong hippocampal marker and Ttr is a marker of the choroid plexus.
samples <- c("Anterior", "Posterior")
to_plot <- c("Hpca", "Ttr")
plist <- list()
n = 1
for (j in to_plot) {
for (i in samples) {
temp <- sce[, sce$Sample == i]
temp@metadata <- temp@metadata[[i]]
plist[[n]] <- spanielPlot(object = temp, plotType = "Gene", gene = j, customTitle = j,
techType = "Visium", ptSizeMax = 1, ptSizeMin = 0.1)
n <- n + 1
}
}
plot_grid(ncol = 2, plotlist = plist)
Dimensionality reduction and clustering
We can then now run dimensionality reduction and clustering using the same workflow as we use for scRNA-seq analysis.
But make sure you run it on the SCT assay.
var.out <- modelGeneVar(sce, method = "loess")
hvgs = getTopHVGs(var.out, n = 2000)
sce <- runPCA(sce, exprs_values = "logcounts", subset_row = hvgs, ncomponents = 50,
ntop = 100, scale = T)
g <- buildSNNGraph(sce, k = 5, use.dimred = "PCA")
sce$louvain_SNNk5 <- factor(igraph::cluster_louvain(g)$membership)
sce <- runUMAP(sce, dimred = "PCA", n_dimred = 50, ncomponents = 2, min_dist = 0.1,
spread = 0.3, metric = "correlation", name = "UMAP_on_PCA")We can then plot clusters onto umap or onto the tissue section.
samples <- c("Anterior", "Posterior")
to_plot <- c("louvain_SNNk5")
plist <- list()
n = 1
for (j in to_plot) {
for (i in samples) {
temp <- sce[, sce$Sample == i]
temp@metadata <- temp@metadata[[i]]
plist[[n]] <- spanielPlot(object = temp, plotType = "Cluster", clusterRes = j,
customTitle = j, techType = "Visium", ptSizeMax = 1, ptSizeMin = 0.1)
n <- n + 1
}
}
plist[[3]] <- plotReducedDim(sce, dimred = "UMAP_on_PCA", colour_by = "louvain_SNNk5")
plist[[4]] <- plotReducedDim(sce, dimred = "UMAP_on_PCA", colour_by = "Sample")
plot_grid(ncol = 2, plotlist = plist)
Integration
Quite often there are strong batch effects between different ST sections, so it may be a good idea to integrate the data across sections.
We will do a similar integration as in the Data Integration lab.
mnn_out <- batchelor::fastMNN(sce, subset.row = hvgs, batch = factor(sce$Sample),
k = 20, d = 50)
reducedDim(sce, "MNN") <- reducedDim(mnn_out, "corrected")
rm(mnn_out)
gc()## used (Mb) gc trigger (Mb) max used (Mb)
## Ncells 11463709 612.3 20655705 1103.2 20655705 1103.2
## Vcells 900767995 6872.4 1537221656 11728.1 1322374276 10089.0Then we run dimensionality reduction and clustering as before.
g <- buildSNNGraph(sce, k = 5, use.dimred = "MNN")
sce$louvain_SNNk5 <- factor(igraph::cluster_louvain(g)$membership)
sce <- runUMAP(sce, dimred = "MNN", n_dimred = 50, ncomponents = 2, min_dist = 0.1,
spread = 0.3, metric = "correlation", name = "UMAP_on_MNN")samples <- c("Anterior", "Posterior")
to_plot <- c("louvain_SNNk5")
plist <- list()
n = 1
for (j in to_plot) {
for (i in samples) {
temp <- sce[, sce$Sample == i]
temp@metadata <- temp@metadata[[i]]
plist[[n]] <- spanielPlot(object = temp, plotType = "Cluster", clusterRes = j,
customTitle = j, techType = "Visium", ptSizeMax = 1, ptSizeMin = 0.1)
n <- n + 1
}
}
plist[[3]] <- plotReducedDim(sce, dimred = "UMAP_on_MNN", colour_by = "louvain_SNNk5")
plist[[4]] <- plotReducedDim(sce, dimred = "UMAP_on_MNN", colour_by = "Sample")
plot_grid(ncol = 2, plotlist = plist)
Do you see any differences between the integrated and non-integrated clusering? Judge for yourself, which of the clusterings do you think looks best? As a reference, you can compare to brain regions in the Allen brain atlas.
Identification of Spatially Variable Features
There are two main workflows to identify molecular features that correlate with spatial location within a tissue. The first is to perform differential expression based on spatially distinct clusters, the other is to find features that are have spatial patterning without taking clusters or spatial annotation into account.
First, we will do differential expression between clusters just as we did for the scRNAseq data before.
# differential expression between cluster 4 and cluster 6
cell_selection <- sce[, sce$louvain_SNNk5 %in% c(4, 6)]
cell_selection$louvain_SNNk5 <- factor(cell_selection$louvain_SNNk5)
markers_genes <- scran::findMarkers(x = cell_selection, groups = cell_selection$louvain_SNNk5,
lfc = 0.25, pval.type = "all", direction = "up")
# List of dataFrames with the results for each cluster
top5_cell_selection <- lapply(names(markers_genes), function(x) {
temp <- markers_genes[[x]][1:5, 1:2]
temp$gene <- rownames(markers_genes[[x]])[1:5]
temp$cluster <- x
return(temp)
})
top5_cell_selection <- as_tibble(do.call(rbind, top5_cell_selection))
top5_cell_selection# plot top markers
samples <- c("Anterior", "Posterior")
to_plot <- top5_cell_selection$gene[1:5]
plist <- list()
n = 1
for (j in to_plot) {
for (i in samples) {
temp <- sce[, sce$Sample == i]
temp@metadata <- temp@metadata[[i]]
plist[[n]] <- spanielPlot(object = temp, plotType = "Gene", gene = j, customTitle = j,
techType = "Visium", ptSizeMax = 1, ptSizeMin = 0.1)
n <- n + 1
}
}
plot_grid(ncol = 2, plotlist = plist)
Single cell data
We can use a scRNA-seq dataset as a referenced to predict the proportion of different celltypes in the Visium spots.
Keep in mind that it is important to have a reference that contains all the celltypes you expect to find in your spots. Ideally it should be a scRNAseq reference from the exact same tissue.
We will use a reference scRNA-seq dataset of ~14,000 adult mouse cortical cell taxonomy from the Allen Institute, generated with the SMART-Seq2 protocol.
First dowload the seurat data from: https://www.dropbox.com/s/cuowvm4vrf65pvq/allen_cortex.rds?dl=1 to folder data/spatial/ with command:
FILE="./data/spatial/allen_cortex.rds"
if [ -e $FILE ]
then
echo "File $FILE is downloaded."
else
echo "Downloading $FILE"
mkdir -p data/spatial
wget -O data/spatial/allen_cortex.rds https://www.dropbox.com/s/cuowvm4vrf65pvq/allen_cortex.rds?dl=1
fi## File ./data/spatial/allen_cortex.rds is downloaded.For speed, and for a more fair comparison of the celltypes, we will subsample all celltypes to a maximum of 200 cells per class (subclass).
allen_reference <- readRDS("data/spatial/allen_cortex.rds")
allen_reference_sce <- Seurat::as.SingleCellExperiment(allen_reference)
# check number of cells per subclass
allen_reference_sce$subclass <- sub("/", "_", sub(" ", "_", allen_reference_sce$subclass))
table(allen_reference_sce$subclass)##
## Astro CR Endo L2_3_IT L4 L5_IT L5_PT
## 368 7 94 982 1401 880 544
## L6_CT L6_IT L6b Lamp5 Macrophage Meis2 NP
## 960 1872 358 1122 51 45 362
## Oligo Peri Pvalb Serpinf1 SMC Sncg Sst
## 91 32 1337 27 55 125 1741
## Vip VLMC
## 1728 67# select 20 cells per subclass, fist set subclass ass active.ident
subset_cells <- lapply(unique(allen_reference_sce$subclass), function(x) {
if (sum(allen_reference_sce$subclass == x) > 20) {
temp <- sample(colnames(allen_reference_sce)[allen_reference_sce$subclass ==
x], size = 20)
} else {
temp <- colnames(allen_reference_sce)[allen_reference_sce$subclass == x]
}
})
allen_reference_sce <- allen_reference_sce[, unlist(subset_cells)]
# check again number of cells per subclass
table(allen_reference_sce$subclass)##
## Astro CR Endo L2_3_IT L4 L5_IT L5_PT
## 20 7 20 20 20 20 20
## L6_CT L6_IT L6b Lamp5 Macrophage Meis2 NP
## 20 20 20 20 20 20 20
## Oligo Peri Pvalb Serpinf1 SMC Sncg Sst
## 20 20 20 20 20 20 20
## Vip VLMC
## 20 20Then run normalization and dimensionality reduction.
allen_reference_sce <- computeSumFactors(allen_reference_sce, sizes = c(20, 40, 60,
80))
allen_reference_sce <- logNormCounts(allen_reference_sce)
allen.var.out <- modelGeneVar(allen_reference_sce, method = "loess")
allen.hvgs = getTopHVGs(allen.var.out, n = 2000)Subset ST for cortex
Since the scRNAseq dataset was generated from the mouse cortex, we will subset the visium dataset in order to select mainly the spots part of the cortex. Note that the integration can also be performed on the whole brain slice, but it would give rise to false positive cell type assignments and and therefore it should be interpreted with more care.
Integrate with scRNAseq
Here, will use SingleR for prediciting which cell types are present in the dataset.
We can first select the anterior part as an example (to speed up predictions).
sce.anterior <- sce[, sce$Sample == "Anterior"]
sce.anterior@metadata <- sce.anterior@metadata[["Anterior"]]Next, we select the highly variable genes that are present in both datasets.
# Find common highly variable genes
common_hvgs <- allen.hvgs[allen.hvgs %in% hvgs]
# Predict cell classes
pred.grun <- SingleR(test = sce.anterior[common_hvgs, ], ref = allen_reference_sce[common_hvgs,
], labels = allen_reference_sce$subclass)
# Transfer the classes to the SCE object
sce.anterior$cell_prediction <- pred.grun$labels
sce.anterior@colData <- cbind(sce.anterior@colData, as.data.frame.matrix(table(list(1:ncol(sce.anterior),
sce.anterior$cell_prediction))))Then we can plot the predicted cell populations back to tissue.
# Plot cell predictions
spanielPlot(object = sce.anterior, plotType = "Cluster", clusterRes = "cell_prediction",
customTitle = "cell_prediction", techType = "Visium", ptSizeMax = 1, ptSizeMin = 0.1)
plist <- list()
n = 1
for (i in c("L2_3_IT", "L4", "L5_IT", "L6_IT")) {
plist[[n]] <- spanielPlot(object = sce.anterior, plotType = "Cluster", clusterRes = i,
customTitle = i, techType = "Visium", ptSize = 0.3, ptSizeMax = 1, ptSizeMin = 0.1)
n <- n + 1
}
plot_grid(ncol = 2, plotlist = plist)
Keep in mind, that the scores are “just” prediction scores, and do not correspond to proportion of cells that are of a certain celltype or similar. It mainly tell you that gene expression in a certain spot is hihgly similar/dissimilar to gene expression of a celltype.
If we look at the scores, we see that some spots got really clear predictions by celltype, while others did not have high scores for any of the celltypes.
We can also plot the gene expression and add filters together, too:
spanielPlot(object = sce.anterior, plotType = "Gene", gene = "Wfs1", showFilter = sce.anterior$L4,
customTitle = "", techType = "Visium", ptSize = 0, ptSizeMin = -0.3, ptSizeMax = 1)
Session info
sessionInfo()## R version 4.1.2 (2021-11-01)
## Platform: x86_64-apple-darwin13.4.0 (64-bit)
## Running under: macOS Catalina 10.15.7
##
## Matrix products: default
## BLAS/LAPACK: /Users/asbj/miniconda3/envs/scRNAseq2022_tmp/lib/libopenblasp-r0.3.18.dylib
##
## locale:
## [1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8
##
## attached base packages:
## [1] stats4 stats graphics grDevices utils datasets methods
## [8] base
##
## other attached packages:
## [1] patchwork_1.1.1 SingleR_1.8.0
## [3] Matrix_1.4-0 biomaRt_2.50.0
## [5] Spaniel_1.2.0 caret_6.0-90
## [7] lattice_0.20-45 SeuratObject_4.0.4
## [9] Seurat_4.0.6 scmap_1.16.0
## [11] scPred_1.9.2 fgsea_1.20.0
## [13] msigdbr_7.4.1 enrichR_3.0
## [15] dplyr_1.0.7 igraph_1.2.11
## [17] pheatmap_1.0.12 clustree_0.4.4
## [19] ggraph_2.0.5 reticulate_1.22
## [21] harmony_1.0 Rcpp_1.0.8
## [23] umap_0.2.7.0 rafalib_1.0.0
## [25] scDblFinder_1.8.0 DoubletFinder_2.0.3
## [27] org.Hs.eg.db_3.14.0 AnnotationDbi_1.56.1
## [29] cowplot_1.1.1 scran_1.22.0
## [31] scater_1.22.0 ggplot2_3.3.5
## [33] scuttle_1.4.0 SingleCellExperiment_1.16.0
## [35] SummarizedExperiment_1.24.0 Biobase_2.54.0
## [37] GenomicRanges_1.46.0 GenomeInfoDb_1.30.0
## [39] IRanges_2.28.0 S4Vectors_0.32.0
## [41] BiocGenerics_0.40.0 MatrixGenerics_1.6.0
## [43] matrixStats_0.61.0 RJSONIO_1.3-1.6
## [45] optparse_1.7.1
##
## loaded via a namespace (and not attached):
## [1] rsvd_1.0.5 ica_1.0-2
## [3] class_7.3-20 ps_1.6.0
## [5] foreach_1.5.1 lmtest_0.9-39
## [7] rprojroot_2.0.2 crayon_1.4.2
## [9] rhdf5filters_1.6.0 spatstat.core_2.3-2
## [11] MASS_7.3-55 nlme_3.1-155
## [13] backports_1.4.1 rlang_0.4.12
## [15] XVector_0.34.0 ROCR_1.0-11
## [17] irlba_2.3.5 callr_3.7.0
## [19] limma_3.50.0 filelock_1.0.2
## [21] xgboost_1.5.0.1 BiocParallel_1.28.0
## [23] rjson_0.2.21 bit64_4.0.5
## [25] glue_1.6.0 sctransform_0.3.3
## [27] parallel_4.1.2 processx_3.5.2
## [29] vipor_0.4.5 spatstat.sparse_2.1-0
## [31] spatstat.geom_2.3-1 tidyselect_1.1.1
## [33] usethis_2.1.5 fitdistrplus_1.1-6
## [35] XML_3.99-0.8 tidyr_1.1.4
## [37] zoo_1.8-9 xtable_1.8-4
## [39] magrittr_2.0.1 evaluate_0.14
## [41] cli_3.1.0 zlibbioc_1.40.0
## [43] miniUI_0.1.1.1 bslib_0.3.1
## [45] rpart_4.1-15 fastmatch_1.1-3
## [47] shiny_1.7.1 BiocSingular_1.10.0
## [49] xfun_0.29 askpass_1.1
## [51] pkgbuild_1.3.1 cluster_2.1.2
## [53] tidygraph_1.2.0 KEGGREST_1.34.0
## [55] tibble_3.1.6 ggrepel_0.9.1
## [57] listenv_0.8.0 Biostrings_2.62.0
## [59] png_0.1-7 future_1.23.0
## [61] ipred_0.9-12 withr_2.4.3
## [63] bitops_1.0-7 ggforce_0.3.3
## [65] plyr_1.8.6 e1071_1.7-9
## [67] dqrng_0.3.0 pROC_1.18.0
## [69] pillar_1.6.4 cachem_1.0.6
## [71] fs_1.5.2 kernlab_0.9-29
## [73] hdf5r_1.3.5 googleVis_0.6.11
## [75] DelayedMatrixStats_1.16.0 vctrs_0.3.8
## [77] ellipsis_0.3.2 generics_0.1.1
## [79] lava_1.6.10 devtools_2.4.3
## [81] tools_4.1.2 beeswarm_0.4.0
## [83] munsell_0.5.0 tweenr_1.0.2
## [85] proxy_0.4-26 DelayedArray_0.20.0
## [87] fastmap_1.1.0 compiler_4.1.2
## [89] pkgload_1.2.4 abind_1.4-5
## [91] httpuv_1.6.5 sessioninfo_1.2.2
## [93] plotly_4.10.0 GenomeInfoDbData_1.2.7
## [95] prodlim_2019.11.13 gridExtra_2.3
## [97] edgeR_3.36.0 deldir_1.0-6
## [99] utf8_1.2.2 later_1.2.0
## [101] BiocFileCache_2.2.0 recipes_0.1.17
## [103] jsonlite_1.7.2 scales_1.1.1
## [105] ScaledMatrix_1.2.0 pbapply_1.5-0
## [107] sparseMatrixStats_1.6.0 lazyeval_0.2.2
## [109] promises_1.2.0.1 R.utils_2.11.0
## [111] goftest_1.2-3 spatstat.utils_2.3-0
## [113] checkmate_2.0.0 rmarkdown_2.11
## [115] statmod_1.4.36 Rtsne_0.15
## [117] uwot_0.1.11 HDF5Array_1.22.0
## [119] survival_3.2-13 ResidualMatrix_1.4.0
## [121] yaml_2.2.1 htmltools_0.5.2
## [123] memoise_2.0.1 locfit_1.5-9.4
## [125] graphlayouts_0.8.0 here_1.0.1
## [127] viridisLite_0.4.0 digest_0.6.29
## [129] assertthat_0.2.1 mime_0.12
## [131] rappdirs_0.3.3 RSQLite_2.2.8
## [133] future.apply_1.8.1 remotes_2.4.2
## [135] data.table_1.14.2 blob_1.2.2
## [137] R.oo_1.24.0 splines_4.1.2
## [139] labeling_0.4.2 Rhdf5lib_1.16.0
## [141] RCurl_1.98-1.5 hms_1.1.1
## [143] rhdf5_2.38.0 DropletUtils_1.14.0
## [145] colorspace_2.0-2 BiocManager_1.30.16
## [147] ggbeeswarm_0.6.0 nnet_7.3-17
## [149] sass_0.4.0 RANN_2.6.1
## [151] fansi_1.0.0 parallelly_1.30.0
## [153] ModelMetrics_1.2.2.2 R6_2.5.1
## [155] grid_4.1.2 ggridges_0.5.3
## [157] lifecycle_1.0.1 formatR_1.11
## [159] bluster_1.4.0 curl_4.3.2
## [161] leiden_0.3.9 testthat_3.1.1
## [163] getopt_1.20.3 jquerylib_0.1.4
## [165] desc_1.4.0 RcppAnnoy_0.0.19
## [167] RColorBrewer_1.1-2 iterators_1.0.13
## [169] stringr_1.4.0 gower_0.2.2
## [171] htmlwidgets_1.5.4 beachmat_2.10.0
## [173] polyclip_1.10-0 purrr_0.3.4
## [175] mgcv_1.8-38 globals_0.14.0
## [177] openssl_1.4.6 batchelor_1.10.0
## [179] codetools_0.2-18 lubridate_1.8.0
## [181] FNN_1.1.3 metapod_1.2.0
## [183] randomForest_4.6-14 prettyunits_1.1.1
## [185] dbplyr_2.1.1 R.methodsS3_1.8.1
## [187] RSpectra_0.16-0 gtable_0.3.0
## [189] DBI_1.1.2 tensor_1.5
## [191] httr_1.4.2 highr_0.9
## [193] KernSmooth_2.23-20 stringi_1.7.6
## [195] progress_1.2.2 reshape2_1.4.4
## [197] farver_2.1.0 viridis_0.6.2
## [199] timeDate_3043.102 xml2_1.3.3
## [201] BiocNeighbors_1.12.0 scattermore_0.7
## [203] bit_4.0.4 spatstat.data_2.1-2
## [205] pkgconfig_2.0.3 babelgene_21.4
## [207] knitr_1.37